Automate Superchain Registry chain support #8105
Conversation
- Converter - Downloader - Lister
- Add compression - Add arg parsing
- We don't want to recurse
- Genesis and configs
emlautarom1
requested review from
flcl42,
deffrian,
stdevMac,
kamilchodola and
jmederosalvarado
|
@stdevMac Can you take care of checking this and making sure it works accordingly with Sedge? Also add to our CI BUT limit those side chains to trigger once per week. |
- Use `0x` when appropriate only
|
@kamilchodola @stdevMac Do not test this branch yet since it's not working ATM. I'll change the status to "ready to review" once I figure out what is the issue (currently Nethermind can't find peers). |
- Fix `author` and `stateUnavailable`
|
@kamilchodola @stdevMac @deffrian Ready to review. The issue was in the |
| return nethermind | ||
|
|
||
|
|
||
| def to_nethermind_runner(chain_name, l1, chain): |
There was a problem hiding this comment.
The WIP OptimismCL section is not included at the moment.
|
|
||
| with open(path.join(tmp_dir, file), "rb") as json_config: | ||
| samples.append(json_config.read()) | ||
| nethermind_dict = zstd.train_dictionary(2**16, samples, threads=-1) |
There was a problem hiding this comment.
The dict_size=2**16 parameter was decided purely by trial-and-error: this value provided the best result.
| } | ||
|
|
||
| using var decompressedStream = new DecompressionStream(streamData); | ||
| decompressedStream.LoadDictionary(buffer.ToArray()); |
There was a problem hiding this comment.
Unfortunately LoadDictionary does not accept any async/Stream overload.
In practice it should not be an issue since the dictionary is ~3.5 Mb and it's only loaded once during initialization.
| @@ -0,0 +1,6 @@ | |||
| #!/usr/bin/env bash | |||
There was a problem hiding this comment.
Add the missing file header:
| #!/usr/bin/env bash | |
| #!/bin/bash | |
| # SPDX-FileCopyrightText: 2025 Demerzel Solutions Limited | |
| # SPDX-License-Identifier: LGPL-3.0-only |
| @@ -0,0 +1,338 @@ | |||
| import argparse | |||
There was a problem hiding this comment.
Add the missing file header:
| import argparse | |
| # SPDX-FileCopyrightText: 2025 Demerzel Solutions Limited | |
| # SPDX-License-Identifier: LGPL-3.0-only | |
| import argparse |
Solves #8065
Changes
superchain.pyscript to fetch, process and generate Nethermind compatible files for all OP Sperchain chains (see https://github.com/ethereum-optimism/superchain-registry)Types of changes
What types of changes does your code introduce?
Testing
Requires testing
If yes, did you write tests?
Notes on testing
Added tests to verify that we're properly decompressing generated
ChainSpecfiles. Some tests were updated to use a newChainSpecFileLoaderthat is format-aware of ChainSpec files, thus is able to decompress them if needed.Documentation
Requires documentation update
Requires explanation in Release Notes
Once this feature is properly working we need to document that Nethermind now supports all OP Superchain chains, not only a small subset (currently: OP, Base and Worldchain).
Remarks
The generated
ChainSpecfiles are stored in a compressed format due to their excessive size when in.jsonformat. Just like in the Superchain Registry repository (https://github.com/ethereum-optimism/superchain-registry/blob/cb9a0ca8eda1608bf9958137a736fd2c18884fbd/superchain/README.md) we use zstandard as the compression algorithm. This is due to the OP team already figuring out that this is one of the best algorithms for this use case, and also to reduce the required amount of code (our scripts already deal with decompression so might as well compress the resulting artifacts using the same algorithm).Currently, 45 configs are generated: 28 in Mainnet and 17 in Sepolia.
We're still missing a GitHub action that automatically runs this script either periodically or when the Superchain Registry repository gets updated (the latter being preferable)